A noval experimental approach will be employed to determine the optimal hematocrit for oxygenation myocardium. This approach avoids the changes in whole body hematocrit which complicated interpretation of previous results because of concurrent changes in myocardial work demand. Experiments will be conducted in the anesthetized, open-chest dog whose left anterior descending coronary artery (LAD) is perfused selectively via an extracorporeal circuit with blood of variable hematocrit (80%-10%), while its remainder, including most of the heart, receives blood with normal hematocrit. In the LAD bed, measurements of total coronary (electromagnetic flowmeter) and regional myocardial blood flows (radioactive microspheres), myocardial oxygen consumption (Fick principle) and contractility (ultrasound), and regional myocardial small vessel hematocrit and small vessel blood volume (indicatory-dilution) will be obtained. In initial studies LAD hematocrit will be varied with perfusion pressure maintained at mean aortic pressure with 1)normal coronary tone, 2) maximal coronary vasocilation (adenosine), and 3) metabolic vasodilation (isoproteronol). In later studies LAD hematocrit will be varied with perfusion pressure reduced by 50% (coronary insufficiency) and during selective, myocardial hypoxia produced by interposing an isolated dog lung in the proposed study will better understanding of basic mechanisms underlying oxygen transport in myocardium. Further, it will provide a basis for selecting fluid therapy in the human patient whose myocardial oxygenation is inadequate because of coronary insufficiency or arterial hypoxia.